See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8638395

Molecular phylogeny and biogeography of the dung beetle genus

Temnoplectron Westwood (Scarabaeidae: Scarabaeinae) from Australia's Wet
Tropics

Article  in  Molecular Phylogenetics and Evolution · June 2004

DOI: 10.1016/j.ympev.2003.09.010 · Source: PubMed

CITATIONS READS

41 204

4 authors, including:

Karen Leanne Bell David K Yeates

University of Western Australia The Commonwealth Scientific and Industrial Research Organisation
40 PUBLICATIONS   318 CITATIONS    309 PUBLICATIONS   6,722 CITATIONS   

SEE PROFILE SEE PROFILE

Craig Moritz
Australian National University
629 PUBLICATIONS   37,809 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

pollen DNA metabarcoding View project

Grinnell Resurvey View project

All content following this page was uploaded by Karen Leanne Bell on 29 May 2018.

The user has requested enhancement of the downloaded file.

 MOLECULAR
PHYLOGENETICS
AND
EVOLUTION
Molecular Phylogenetics and Evolution 31 (2004) 741–753
www.elsevier.com/locate/ympev

Molecular phylogeny and biogeography of the dung beetle genus

Temnoplectron Westwood (Scarabaeidae: Scarabaeinae)
from AustraliaÕs wet tropics
K.L. Bell,a,b,* D.K. Yeates,a,c C. Moritz,a,b,d and G.B. Monteithe
a
Department of Zoology and Entomology, School of Life Sciences, University of Queensland, St Lucia, Queensland 4072, Australia
b
Rainforest CRC, School of Life Sciences, University of Queensland, St Lucia, Queensland 4072, Australia
c
Australian National Insect Collection, CSIRO Entomology, PO Box 1700, Canberra, ACT 2601, Australia
d
Museum of Vertebrate Zoology, Department of Integrative Biology, 3101 Valley Life Sciences Building #3160, UC Berkeley, CA 94720-3160, USA
e
Queensland Museum, Brisbane, Queensland 4000, Australia

Received 17 June 2003; revised 8 September 2003

Abstract

The landscape of the Australian Wet Tropics can be described as ‘‘islands’’ of montane rainforest surrounded by warmer or more
xeric habitats. Historical glaciation cycles have caused expansion and contraction of these rainforest ‘‘islands’’ leading to consistent
patterns of genetic divergence within species of vertebrates. To explore whether this dynamic history has promoted speciation in
endemic and diverse groups of insects, we used a combination of mtDNA sequencing and morphological characters to estimate
relationships and the tempo of divergence among Australian representatives of the dung beetle genus Temnoplectron. This phylo-
genetic hypothesis shares a number of well-supported clades with a previously published phylogenetic hypothesis based on mor-
phological data, though statistical support for several nodes is weak. Sister species relationships well-supported in both tree
topologies, and a tree obtained by combining the two data sets, suggest that speciation has mostly been allopatric. We identify a
number of speciation barriers, which coincide with phylogeographic breaks found in vertebrate species. Large sequence divergences
between species emphasize that speciation events are ancient (pre-Pleistocene). The flightless, rainforest species appear to have
speciated rapidly, but also in the distant past.
Ó 2003 Elsevier Inc. All rights reserved.

Keywords: Biogeography; Speciation; Molecular phylogeny; Tropical rainforest; Dung beetle

1. Introduction of this concept may be circular) and niche exploitation

The biodiversity of tropical rainforests has fascinated
biologists since at least the 19th century. The species
richness of these rainforests is higher than that of any
other ecosystem (Fischer, 1960). The high level of bio-
diversity in tropical rainforest has been explained with
both ecological and historical hypotheses. Ecological
factors that could promote species richness include
habitat productivity (e.g., Givnish, 1999), habitat het-
erogeneity (e.g., Kikkawa et al., 1981), habitat ‘‘fa-
vourableness’’ (e.g., Brown, 1988, although definitions
*
Corresponding author. Fax: +61-7-33651655.
E-mail address: kbell@zen.uq.edu.au (K.L. Bell).
(e.g., Farrell, 1998). Historical processes that could lead
to higher species richness in the tropics include rain-
forest contraction and isolation during more arid peri-
ods, and habitat persistence within rainforest refugia at
these times (e.g., Avise et al., 1998; Bush, 1994).
Pliocene–Pleistocene global climatic cycles (for the
last 2.8MY) are thought to have caused the fragmen-
tation of montane habitats, with the potential to drive
allopatric speciation. For rainforest species the most
obvious types of disruption are gaps of dry habitat be-
tween rainforest, although these may not be obvious in
the current landscape because the rainforest is more
extensive and continuous than in previous, drier periods.
While palynological evidence for rainforest contraction
in most tropical regions is ambiguous (Colinvaux, 1987),

1055-7903/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.ympev.2003.09.010
742 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753
pollen cores from the Australian Wet Tropics (hereafter,
the ‘‘Wet Tropics’’; the rainforest region of northeast
Queensland between Townsville and Cooktown) reveal
a clear history of severe rainforest contraction during
Pleistocene glacial periods (Kershaw, 1994). These data
are consistent with palaeoclimate models, which predict
that temperature and moisture regimes suitable for
rainforest growth would have been fragmented and
disjunct throughout the region (Nix, 1991).
Phylogeographic analyses of Wet Tropics vertebrate
species strongly suggest refugial isolation and allopatric
molecular divergence, but not complete speciation, across
barriers (e.g. Schneider and Moritz, 1999; Schneider et al.,
1999). The magnitude of molecular divergence varies
among species, possibly reflecting more than one expan-
sion and contraction, mostly pre-Pleistocene, across these
boundaries (Joseph and Moritz, 1993; Joseph et al., 1995;
Schneider et al., 1998). There have, however, been very
few studies on the biogeography of endemic radiations in
the Wet Tropics with which to explore the effects of
rainforest contractions on speciation.
By studying species-rich and taxonomically well-
known groups in rainforests, we can gain insight into the
origins of the high level of biodiversity, and the causes of
speciation. Dung beetles (Coleoptera: Scarabaeidae:
Scarabaeinae) provide an ideal system for studying en-
demic radiations in the Wet Tropics. The Australian
Scarabaeinae has been comprehensively reviewed, so
that the majority of species are described, with special
focus on the fauna of northeastern Queensland (Mat-
thews, 1972; Matthews, 1974; Matthews, 1976). Since
they occupy smaller ranges, invertebrates tend to be
informative at a finer scale than are vertebrates (Moritz
et al., 2001), and studies of invertebrates have shown
evidence for vicariant speciation between Wet Tropics
rainforest blocks, rather than just molecular divergence
(e.g., Baehr, 1995; Bouchard et al., in press; Reid and
Storey, 2000). Dung beetles can be collected efficiently in
baited traps, and continuous and extensive collecting
means that the distributions are known for most de-
scribed species. Many species of Australian dung beetle
are recorded as being habitat specific (Matthews, 1972;
Matthews, 1974; Matthews, 1976). There is a distinct
rainforest and open forest fauna, (Hill, 1996).
The genus Temnoplectron has recently been revised
taxonomically (Reid and Storey, 2000). Its species are
predominantly Wet Tropics endemics, and it has radi-
ated within the Wet Tropics. The genus contains 19
species, 11 of which are restricted to a portion of the
Wet Tropics (Reid and Storey, 2000, Fig. 1). These
narrow endemics are commonly allopatric or parapatric
with respect to their sister taxa according to the clado-
gram based on morphological characters (Reid and
Storey, 2000). Many of these sister taxa have strongly
divergent genitalia, suggesting reproductive isolation,
with very little divergence in external morphology and
may be useful for studying the effects of rainforest
fragmentation on speciation. The genus also has the
potential to answer questions relating to the effect of
flight loss on speciation, as it contains both volant and
non-volant species (Reid and Storey, 2000).
Flightlessness leads to a decrease in mobility and,
hence, it would be expected to increase the potential for
allopatric or parapatric speciation (Roff, 1994). Several
lines of evidence support this hypothesis; in general,
flightless species are more concentrated in refugia, are
found at higher elevations and have higher levels of
endemism. The ranges of flightless carabid beetles are
significantly smaller than those of volant species, and
gene flow among flightless populations is less than
among populations of volant species (Liebherr, 1986).
For the Wet Tropics, Yeates et al. (2002) found high
levels of endemism in flightless insects, although this was
not directly compared to volant insect species. There
have, however, been no studies of speciation rates or
studies of gene flow within populations of volant and
non-volant insects in this geographic region. The phy-
logenetic hypothesis of Reid and Storey (2000) suggests
that flightlessness in Temnoplectron has led to a rela-
tively high level of speciation within a small geographic
range. Thus it would be expected that these flightless
Temnoplectron species have speciated more rapidly than
the fully winged species in this genus. A rapid speciation
rate at some time in the past would lead to closely
spaced cladogenesis events among flightless taxa in a
molecular phylogeny.
2. Materials and methods
2.1. Sampling
All specimens included in this study are shown in
Table 1. For each species at least two individuals, where
possible from geographically distant sites, were used.
Where multiple isolated populations of a species oc-
curred, individuals from all populations were included.
Species from New Guinea were not included, as the
specimens could not be obtained. The species Temno-
plectron diversicolle Blackburn, found at only two lo-
calities, in lowland rainforest, approximately 40 km
north of Cooktown, was also omitted as the most re-
cently collected specimens were from 1976, and no
readable nucleotide sequence was obtained.
2.2. Mitochondrial DNA sequencing
Up to 1330 bp of the mitochondrial gene cytochrome
oxidase I (COI) was sequenced using the primers shown
in Table 2 (GenBank Accession Nos. AY144730–
AY144795). For fresh specimens, stored in 100% etha-
nol, the PCR product was obtained in one fragment.
 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753 743

Fig. 1. Distribution data for Temnoplectron species. (A) Listing of subregions in which each species is present. (B) Map showing subregions within the
Wet Tropics. After Williams et al. (1996). CL, Cooktown Lowlands; FU, Mount Finnigan Uplands; TU, Thornton Uplands; WU, Windsor Uplands;
CU, Carbine Uplands; BM, Black Mountain Corridor; LU, Lamb Uplands; AU, Atherton Uplands; BK, Bellenden-Ker/Bartle-Frere; MT, Malbon-
Thompson Uplands; KU, Kirrama Uplands; LE, Lee Uplands; HI, Hinchinbrook Island; SU, Spec Uplands; HU, Halifax Uplands; EU, Elliot
Uplands; PI, Palm Islands (C) map showing subregions outside the Wet Tropics. NG, New Guinea; NWA, North Western Australia; NT, Northern
Territory; CY, Cape York; MEQ, Mid-East Queensland; SEQ, South-East Queensland.

For older or poorly preserved museum specimens, in-
ternal primers were used to obtain 150, 300 or 500 bp
products. For all species, multiple individuals (up to 5)
were sequenced. The first 436 bp were consistently ob-
tained, but longer sequence of COI (>800 bp) was usu-
ally obtained for only one individual. All amplifications
used standard PCR conditions followed by automated
sequencing using Applied Biosystems Big-Dye termina-
tor sequencing chemistry. Sequences were edited in Se-
quence Navigator, and aligned using ClustalX.
2.3. Phylogenetic analysis
Aligned sequences were analysed using PAUP
ver-
sion 4.0b10. Firstly, a maximum likelihood tree was
obtained using the first 436 bp of COI sequence obtained
for all specimens in Table 1, using Monoplistes Lans-
berge, Diorygopyx Matthews, and Lepanus Balthasar as
outgroups. The purpose of this was to test the mono-
phyly of all species, as defined by Reid and Storey
(2000). Parameters for maximum likelihood analysis
using a model of general time reversible plus gamma
distribution (GTR + C) were estimated based on pre-
liminary topologies obtained by neighbour joining. A
maximum likelihood tree was obtained by a heuristic
search using a branch-swapping algorithm of subtree-
pruning-regrafting (SPR). Bayesian support values for
nodes on this tree were also obtained using MrBayes
2.01. If any species were not monophyletic in the tree
topology, the analysis was repeated constraining those
species to be monophyletic, and the likelihood scores of
the constrained and unconstrained trees were compared
with a SH-test (Shimodaira and Hasegawa, 1999).
Once monophyly of the species was confirmed by the
above analysis, the longest sequence obtained for each
species (408–994 bp, mean 933) was used in maximum
likelihood analysis. The maximum likelihood tree was
obtained by a heuristic search and Bayesian support
values were also obtained using MrBayes 2.01 as de-
scribed above. This phylogenetic hypothesis was tested
for congruence with the cladogram based on morphol-
ogy of Reid and Storey (2000) by producing a ML tree
 744
Table 1
Listing of all specimens for which COI sequence was obtained, and information on their collection
Species Collection locality Altitude Collector(s) Date Specimen code no.
0 0
T. aeneopiceum Matthews Tully R. causeway (17°54 S 145°39 E) 750 m D.J. Cook 2 Dec 1998 2002-1

K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753

Kjellberg Rd (17°320 S 145°350 E) 920 m G.B. Monteith 6–9 Feb 1999 2154-1
Millaa Millaa Falls (17°290 48:700 S 145°360 34:000 E) 825 m K.L. Bell 8–10 Dec 2000 AT-1
Tully Falls Rd (17°430 4900 S 145°320 1800 E) K.L. Bell 9–10 Jul 2000 WH-1
Birthday Creek Falls (18°580 42:100 S 146°090 34:800 E) 880 m K.L. Bell 1–2 Dec 2000 PR-1
T. bornemisszai Matthews Bartle Frere, W. base (17°230 S 145°460 E) 750 m G.B. Monteith & D.J. Cook 28–30 Nov 1998 1962-1
Tully R. causeway (17°540 S 145°390 E) 750 m D.J. Cook 3–5 Dec 1998 2015-1
Mt Hugh Nelson, S. base (17°310 S 145°330 E) 1080 m G.B. Monteith & D.J. Cook 7–9 Feb 1999 2167-1
T. boucomonti Paulian Station Creek (16°370 S 145°140 E) 350 m G.B. Monteith 23–25 Nov 1998 1938-1
Gregory NP (16°070 S 130°250 E) P. Bouchard Jun 2001 NT-1
T. cooki Reid and Storey Mt Spurgeon (16°270 S 145°110 E) G.B. Monteith, D.J. Cook & C. Burwell 19–23 Nov 1997 1632-1
Mt Haig (17°070 S 145°330 E) 1070 m G.B. Monteith & D.J. Cook 6–10 Feb 1998 1798-1
1798-2
T. disruptum Matthews Iron Range (12°460 S 143°170 E) J. Skevington 20–24 Jul 1999 UQ-1
UQ-2
T. finnigani Reid and Storey Mt Finnigan (15°490 S 145°170 E) 1100 m G.B. Monteith 20–21 Nov 1998 1929-1
1929-2
1929-3
T. involucre Matthews Birthday Creek Falls (18°580 42:100 S 146°090 34:800 E) 880 m K.L. Bell 1–2 Dec 2000 PR-1
Paluma Dam Rd (19°00:7160 S 146°10:4610 E) 660 m K.L. Bell 2–5 Dec 2000 PR-2
PR-4
T. laeve Laporte Cameron Ck, via Koumala (21°350 S 149°140 E) G.B. Monteith 1–2 Oct 1999 7799-1
7799-2
Birthday Creek Falls (18°580 42:100 S 146°090 34:800 E) 880 m K.L. Bell 1–2 Dec 2000 PR-1
Mt Spurgeon (16°270 S 145°110 E) G.B. Monteith, D.J. Cook & C. Burwell 19–23 Nov 1997 T58245 (QM)
T58246 (QM)
T. lewisense Reid and Storey Mt Lewis, site 3 (16°330 04:200 S 145°160 45:600 E) 980 m K.L. Bell 13–14 Dec 2000 CT-1
Mt Lewis barracks (16°350 4000 S 145°160 2700 E) 990 m 19 Apr 1997 QM-1
T. major Paulian Taroom (25°360 S 149°460 E) 200 m D.J. Cook & G.B. Monteith 15–16 Dec 2000 9853-1
Ravenshoe, 2km SW (17°370 S 145°250 E) 880 m G.B. Monteith 6–8 Feb 1999 T72896 (QM)
T. monteithi Reid and Storey Cape Tribulation (16°050 S 145°260 E) G.B. Monteith 17 Nov 1998 1913-1
1913-2
T. politulum Macleay Bartle Frere, W. base (17°230 S 145°460 E) 750 m G.B. Monteith 28 Nov 1998 1970-1
Vine Creek (17°410 S 145°320 E) 1060 m G.B. Monteith 4–6 Feb 1999 2094-1
Wallaman Falls (18°350 26:600 S 145°480 16:300 E) 300 m K.L. Bell 2 Dec 2000 SR-1
T. reyi Paulian Cape Tribulation (16°050 S 145°260 E) G.B. Monteith 17 Nov 1998 1913-1
Mt Lewis, site 1 (16°350 11:900 S 145°170 19:900 E) 860 m K.L. Bell 13–14 Dec 2000 CT-2
2km ENE Kuranda (16°480 S 145°380 E) 360 m C. Reid 19–21 Apr 1999 ANIC-2
Cedar Pocket (16°560 S 145°360 E) 440 m I. Reid & C. Reid 17–19 Apr 1999 ANIC-3
 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753 745

constrained to be consistent with this cladogram. The

likelihoods of the two phylogenetic hypotheses were
T50617 (QM)

T50618 (QM)

compared with a SH-test (Shimodaira and Hasegawa,

1999).
1970-1
2035-1

7819-1
7819-2
7579-1

7579-2
7579-3
7579-4
UQ-1
UQ-2
AT-1
CT-2
SQ-1
To increase support for the tree, the molecular data
were combined with the data set of Reid and Storey
(2000). The combined data set was analysed by parsi-

18 Dec 1998–25 Jan

mony. Molecular and morphological characters were
17–18 Mar 2001
20–24 Jul 1999
8–10 Dec 2000

unordered and equally weighted. Heuristic search set-

7–9 Nov 1999
7–8 Dec 1998
5–8 Feb 1975

28 Nov 1998

11 Jul 2000

tings were as previously used for maximum likelihood

analysis. Partitioned Bremer support (Baker and DeS-
1999

alle, 1997) was applied using TreeRot version 2.0, to

determine if there was any conflict between the two data
sets, and in which regions of the tree conflict occurred.

3. Results
P. Bouchard & A. McDowell

G.B. Monteith & D.J. Cook

K.L. Bell & G.B. Monteith

3.1. Confirmation of current taxonomy

Monteith & Gough

To test for monophyly of species according to the

current definitions (Reid and Storey, 2000), a maximum
G.B. Monteith

G.B. Monteith
G.B. Monteith

J. Skevington

likelihood tree was produced with the first 436 bp of each

K.L. Bell

sequence (Fig. 2). This phylogenetic hypothesis has a

likelihood score of 4653.19535, and confirms the mono-
phyly of all species, with bootstrap values >80%, except
Temnoplectron politulum Macleay, which for mtDNA
appears paraphyletic without including Temnoplectron
750 m

825 m

728 m

580 m

reyi Paulian, its hypothesised sister taxon (Reid and

Storey, 2000). The location of the root within this clade is
uncertain and the analysis was repeated with T. politulum
Sunday Ck barracks, 6km NE (26°400 S 152°350 E)
Millaa Millaa Falls (17°290 48:700 S 145°360 34:000 E)

constrained to be monophyletic. The tree topology ob-

tained had a )log-likelihood of 4655.15683. A SH-test of
Macleay Nipping Gully (25°200 S 151°260 E)

the two tree topologies gave P ¼ 0:257, which is not

Bartle Frere, W. base (17°230 S 145°460 E)
Andoom, nr Weipa (12°320 S 141°500 E)

significant. In addition, DNA sequencing of 96 individ-

Bulburin barracks (24°320 S 151°280 E)
Mt Murray Prior (16°560 S 145°510 E)

Mt Lewis (16°350 1400 S 145°180 0700 E)

uals of T. politulum and T. reyi suggests monophyly of

both species (KLB, CM, A. Moussalli & DKY, in prep).
Iron Range (12°460 S 143°170 E)

Therefore, we treat all species as monophyletic.

3.2. Sequence divergence within and between species

For specimens taken from up to 100 km apart,

within-species sequence divergence (using ML model) in
the winged species ranged from 0 to 9.4% (mean 3.9%).
Sequence divergence within the wingless species is
higher, ranging from 0.23 to 13.2% (mean 6.7%). Esti-
mates of sequence divergence within the wingless species
Lepanus monteithi Matthews

Lepanus ustulatus Lansberge

Diorygopyx tibialis Macleay

can be substantial, even for specimens collected from the

T. subvolitans Matthews

same location.
Monoplistes occidentalis
T. rotundum Westwood

Sequence divergences between species are shown in

Fig. 3. The level of divergence between sister-species for
994 bp of sequence is highly variable, ranging from 6.4%
between T. politulum and T. reyi, to 21.5% between
Macleay

Temnoplectron involucre Matthews and Temnoplectron

monteithi Reid and Storey. The level of net divergence
(i.e., corrected for variation within species) between
746 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753

Table 2
Primers used for sequencing of COI
Designation Alias Sequence Reference
C1-J-1706 CO1-FA CCA ATT TTA ATT GGR GGR TTT GG Custom primer
C1-J-1718 Geoff GGA GGA TTT GGA AAT TGA TTA GTT CC (Simon et al., 1994)
C1-J-1751 Ron GGA TCA CCT GAT ATA GCA TTC CC (Simon et al., 1994)
C1-J-1752 CO1-F5-D GAG CCC CTG ATA TAG CCT TTC C Custom primer
C1-J-1754 Johnny GCT CCT GAT ATA GCA TTC CCA CG Custom primer
C1-J-1844 CO1-150-F GTT GAA AGA GGA GCT GGA ACA GG Custom primer
C1-N-1841 CO1-150-R TGA TAA AGG RGG ATA AAC TGT TC Custom primer
C1-N-1970 MiniCO1 TAT TCC TGT TGA TCG TAT ATT RA Custom primer
C1-J-1994 CO1-300-F CAA TAT ACG ATC AAC AGG AAT AAC Custom primer
C1-J-2166 CO1-F3 GGA GAC CCA ATT CTT TAY CAA C Custom primer
C1-J-2166 C01-F3-D GGA AAC CCG ATT TTA TAT CAA C Custom primer
C1-J-2183 Jerry CAA CAT TTA TTT TGA TTT TTT GG (Simon et al., 1994)
C1-N-2191 Nancy CCC GGT AAA ATT AAA ATA TAA ACT TC (Simon et al., 1994)
C1-J-2195 CO1-RLR TTG ATT TTT TGG TCA TCC AGA AGT (Simon et al., 1994)
C1-N-2329 K525 ACT GTA AAT ATA TGA TGA GCT CA (Simon et al., 1994)
C1-N-2454 CO1-800-R GGG AAT AAT TTA ATT GAG ATC C Custom primer
C1-J-2474 CO1-800-F CTC TTC ATG GAT CTC AAT TAA ATT AYT C Custom primer
C1-N-2652 CO1-R3 TCC TGT AAA TAA TGG AAA TCA ATG Custom primer
C1-N-2751 Jeanette CCA CTT AAT CCT AAA AAA TGY TG Custom primer
C1-N-2751 Pauline CCA CTT AAA CCA AAR AAA TGY TG Custom primer
L2-N-3014 Pat TCC AAT GCA CTA ATC TGC CAT ATT A (Simon et al., 1994)

sister-species for 437 bp ranges from 4.0% between T.
politulum and T. reyi, to 15.2% between T. involucre and
T. monteithi.
3.3. Maximum likelihood analysis of full length COI
sequence
While the tree topology obtained using 436 bp of COI
(Fig. 2) supports many of Reid and StoreyÕs (2000) hy-
potheses of sister-species relationships (Fig. 4B), the tree
tends to have short internal branch lengths and corre-
spondingly low levels of support. To obtain higher
support levels, longer sequences of COI were used. For
most species, we obtained at least a 994 bp sequence, the
exceptions being T. rotundum Westwood (408 bp) and
Temnoplectron lewisense Reid and Storey (781 bp). The
resulting phylogenetic hypothesis (Fig. 4A) has a )log-
likelihood of 5520.55169.
Support for nodes was highly variable, with essen-
tially no support for those at the base, but strong sup-
port, especially with regard to Bayesian posterior
probabilities, for several nodes involving species en-
demic to the Wet Tropics (Fig. 4A). A well-supported
node separates a major clade of northeast Queensland
and mostly rainforest-restricted species from a group of
more ecologically generalized and widely distributed
taxa, among which relationships are poorly defined.
Considering the former clade, the Cape York rainforest
endemic Temnoplectron disruptum Matthews appears to
be outside of a moderately supported clade comprising
nine Wet Tropics endemics, of which all but one
(Temnoplectron cooki Reid and Storey, in wet sclero-
phyll forest) are restricted to rainforest. Within this
clade is a monophyletic group of wingless species. Each
of these rainforest-restricted and wingless species is
found in a single subregion, but area relationships can-
not be deduced because of poor support for nodes
within this clade. Among the four winged rainforest
species from the Wet Tropics, two essentially parapatric
and widely distributed pairs are identified with strong
support; (Temnoplectron aeneopiceum Matthews,
Temnoplectron subvolitans Matthews; 14.2% divergence)
and (T. politulum, T. reyi; 6.4% divergence). The former
is related to the wingless clade whereas the latter is sister
group to the wingless wet-sclerophyll species, T. cooki.
Relatively few nodes on the morphology-based tree
of Reid and Storey (2000, Fig. 4B) were strongly sup-
ported, but there are some areas of consistency with the
mtDNA tree (Fig. 4A) and also some potential conflicts.
In both analyses, the four wingless rainforest species are
monophyletic, as are the winged clades of (T. aeneopi-
ceum, T. subvolitans) and (T. politulum, T. reyi). The
only potential conflict between strongly supported nodes
of the mtDNA tree and the morphological cladogram
concern the relationships of (T. aeneopiceum, T. sub-
volitans); this clade is monophyletic with the wingless
clade in the mtDNA tree, but appears as a sister group
to (T. politulum, T. reyi) and to the exclusion of T. cooki
in the morphological analysis.
Constraining the maximum likelihood tree to have
only clades consistent with the poorly resolved mor-
phological tree of Reid and Storey (2000) produced a
phylogenetic hypothesis with a )log-likelihood of
5536.79051, which is not significantly different to the
unconstrained tree ()log-likelihood of 5520.55169), ac-
cording to a SH-test (Shimodaira and Hasegawa, 1999).
 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753 747

Fig. 2. Maximum likelihood tree based on 437 bp of COI, with bootstrap values shown above branches and Bayesian support values shown below
branches.

Combining the molecular and morphological characters 4. Discussion

in a ‘‘total evidence,’’ unweighted parsimony analysis
(Fig. 5) returned the congruent relationships identified
above and provided additional support within the
wingless clade, with 87% bootstrap support for (T.
lewisense, T. monteithi), species restricted to the adjacent
Carbine and Thornton uplands, respectively. Aside from
these and the congruent relationships (above), bootstrap
support for individual nodes remained low (<80%).
Partitioned Bremer support shows some conflict be-
tween the two datasets. This is predominantly localised
to the basal lineages which also have low support in the
separate analyses of the two data sets.
4.1. Speciation patterns within the Wet Tropics
Overall, the phylogenetic hypothesis produced for
Temnoplectron had variable support. For the tree
produced using mtDNA, support was low at the base,
possibly due to site saturation, but within the Wet
Tropics clade support was higher, except among the
wingless species. In addition, there was some conflict
between the molecular and morphological data sets
in the basal part of the tree produced with the com-
bined data. Where the phylogenetic hypothesis is well
748 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753

Fig. 3. (A) Histogram of total interspecific sequence divergences for 994 bp of COI based on a maximum likelihood model. (B) Histogram of net
interspecific sequence divergences (ie. corrected for variation within species) for 437 bp of COI based on a maximum likelihood model.

supported we can make some inferences about speci-
ation processes.
There are a remarkable number of sister species with
adjacent ranges according to the tree topology produced
with the combined data sets (see Figs. 6–8 for distribu-
tions of these species). The distributions of 47% of
species are non-overlapping with their sister taxa. Vi-
cariance barriers coincide with those observed in other
species (Bouchard et al., in press; Joseph et al., 1995;
Pope et al., 2000; Schneider and Moritz, 1999; Schneider
et al., 1998) and predicted breaks in rainforest during
Pleistocene, and earlier, glaciations (Nix, 1991). There is
evidence for allopatric speciation being a major cause of
speciation in this genus.
Temnoplectron subvolitans is known from two popu-
lations—the northern population is centered on the
Carbine Uplands, while the southern population occurs
on the Atherton Uplands and Lamb Uplands between
Millaa Millaa Falls and Cairns. T. aeneopiceum, sister of
T. subvolitans, is found in rainforest from the Halifax
Uplands north to the southern end of the Atherton
Uplands. The two species are in contact at the southern
end of the Atherton Uplands (Fig. 6). Brachyptery (re-
duced wings) occurs in T. subvolitans in this region of
overlap. No morphological traits expected in hybrids
have been observed (Reid and Storey, 2000). A study by
Bouchard et al. (in press) of arthropod macroevolution
in the Wet Tropics using secondary Brooks Parsimony
Analysis (an approach to historical biogeography which
uses area cladograms, duplicating areas with reticulate
histories, Brooks et al., 2001) included the genus
Temnoplectron, using the cladogram based on morpho-
 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753 749

Fig. 4. (A) Maximum likelihood tree based on 994 bp of COI with bootstrap values shown above branches and Bayesian support values shown below
branches. (B) Consensus of most parsimonious trees based on morphological data from Reid and Storey (2000).

logical characters (Reid and Storey, 2000). Consistent
with our results, they concluded that these two species
were probably involved in a vicariant event, with T.
subvolitans speciating in Bellenden–Ker/Bartle–Frere,
and T. aeneopiceum speciating in the Atherton Uplands,
followed by extensive post speciation dispersal in both
species and probably ecological differentiation. A de-
tailed analysis of the population genetics of T. aeneopi-
ceum and T. subvolitans is underway to further
investigate the speciation processes involved. This hy-
pothesised vicariant speciation event at the southern end
of the Atherton Uplands is coincident with secondary
contacts of phylogroups in widespread vertebrates, in-
cluding the frogs Litoria rheocola and Litoria nannotis
(Schneider et al., 1998), and in a widespread snail
Gnarosophia bellendenkerensis (Hugall et al., 2002).
Temnoplectron reyi is known from the Mount Finni-
gan Uplands area, south to the northern slopes of the
Lamb Uplands and shows little divergence across its
range. ItÕs sister species, T. politulum is found from the
Lamb Uplands south to the Lee Uplands (Fig. 7). Col-
lection records show the two species to have adjacent,
non-overlapping ranges, within 6 km of each other in the
Lamb Uplands. Secondary Brooks Parsimony Analysis
by Bouchard et al. (in press) suggested that these species
evolved by peripheral isolation with an unknown an-
cestor. However, they also concluded that the distribu-
tion pattern could possibly be explained by a vicariant
event with a post-speciation dispersal of T. reyi into the
Lamb Uplands. Our evidence favours the latter hy-
pothesis in that the geographic break across the Lamb
Uplands coincides with secondary contacts of phylo-
groups in widespread vertebrates, including the frog Li-
toria genimaculata (Schneider et al., 1998), the skink
Carlia rubrigularis (B. Phillips, unpublished) and the
small marsupial Bettongia tropica (Pope et al., 2000) and
in a widespread snail G. bellendenkerensis (Hugall et al.,
2002).
Although the molecular, morphological, and com-
bined tree topologies differ in the arrangement of the
clade of wingless, montane, rainforest species, branch
lengths on the molecular tree imply almost simultaneous
speciation of these four species. T. monteithi (found on
Thornton Uplands) and T. finnigani Reid and Storey
750 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753

Fig. 5. Consensus of two most parsimonious trees of combined molecular and morphological data with bootstrap values shown above branches, and
partitioned Bremer support values (molecular/morphological) shown below branches.

(found on Mount Finnigan Uplands) are separated by
the Bloomfield River. Bouchard et al. (in press) hy-
pothesised that these two species evolved through a vi-
cariant event. The valley of the Daintree River separates
T. lewisense (found on the Carbine Uplands) and T.
monteithi (Thornton Uplands). The other species in this
clade, T. involucre, is found in the Spec Uplands and
Halifax Uplands, far to the south of the other species.
Bouchard et al. (in press) suggest that both T. lewisense
and T. involucre are the result of peripheral isolate
speciation, from an ancestor common to all four of the
above species. Reid and Storey (2000) hypothesized
speciation by vicariance rather than dispersal. This ex-
planation was supported by the variation of tibial
morphology, in which the currently geographically and
sexually isolated species appear to preserve the longi-
tudinally clinal range of variation of a widespread an-
cestor (Reid and Storey, 2000). Data from our study is
inconclusive in determining mode of speciation for these
species, as support values are low. However, it does
show that speciation was rapid and ancient in this clade.
It is noteworthy that there are no wingless Temnoplec-
tron species present on the Atherton Uplands or
Bellenden–Ker/Bartle–Frere. Interestingly, there are
species of the wingless scarabaeine genus Aptenocanthon
Matthews present in these areas, which may occupy a
similar ecological niche to wingless Temnoplectron spe-
cies (Storey and Monteith, 2000).
 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753
Fig. 6. Map of the Wet Tropics World Heritage Area, showing
distribution of T. subvolitans (black) and T. aeneopiceum (light grey),
and overlap between species (dark grey).
Three species isolated on either side of known vicar-
iance barriers have failed to diverge morphologically in
characters used to discriminate species, but have di-
verged in size and microsculpture (Reid and Storey,
2000). These species also have low levels of sequence
divergence between the populations, relative to that
found between species. The flightless T. cooki has iso-
lated populations on Carbine Uplands and Lamb Up-
lands, separated by the Black Mountain Corridor, with
a divergence between populations of 8%. Genetic di-
vergence across the Black Mountain Corridor has been
recorded in many Wet Tropics species (Joseph et al.,
1995; Schneider et al., 1999). This break separates the
two main rainforest refugia identified by Nix (1991).
T. subvolitans has populations on the Carbine Uplands
and (Atherton Uplands + Lamb Uplands), also sepa-
rated by the Black Mountain Corridor, with 3% se-
quence divergence. Further studies of population
genetics will more accurately determine this sequence
divergence. Temnoplectron laeve Laporte has four iso-
751
Fig. 7. Map of the Wet Tropics World Heritage Area, showing
distribution of T. politulum (grey) and T. reyi (black).
lated populations—Carbine Uplands, (Spec Up-
lands + Halifax Uplands), Mid-East Queensland and
Palm Islands—one (Carbine Uplands) of which is par-
tially sympatric with its sister species, T. bornemisszai
Matthews. Within species variation in this widely dis-
tributed and winged species is relatively low (max 2%).
4.2. Effect of flight loss on genetic diversity and speciation
rate
There are two origins of flightlessness in this genus—a
monophyletic group of four wingless, montane, rain-
forest species, and an individual species, T. cooki, which
is found in two isolated populations in montane wet
sclerophyll forest. There is also a reduction of wings in
specimens of T. subvolitans in the area where they are in
contact with their sister species, T. aeneopiceum. There is
no indication of recent speciation in the wingless
752 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753
Fig. 8. Map of the Wet Tropics World Heritage Area, showing
distributions of wingless rainforest species.
clade—winged as well as wingless sister taxa are highly
divergent. While we do not have sufficient data to make
precise comparisons between speciation rates in fully
winged and flightless lineages, in the flightless lineage,
short branch lengths of ancestral nodes suggest almost
simultaneous speciation of these species.
Acknowledgments
The authors would like to thank the Queensland
Museum, the Australian National Insect Collection and
the University of Queensland Insect Collection for
providing a substantial proportion of the beetle speci-
mens used in this project. We would also like to thank
Chris Reid for helpful discussion, and for making his
data on the revision and systematics of Temnoplectron
available to us prior to its publication. The authors are
also grateful to Andrew Hugall for his assistance with
laboratory work, and useful suggestions for data anal-
ysis. We thank Gaynor Dolman, Dave Walter, Margaret
Schneider, and two anonymous reviewers for comments
on the manuscript. We also would like to thank Karen
Richardson and Josie Carwardine for providing the
maps used in the figures. This project was funded by the
Rainforest CRC research support scheme, and an Aus-
tralian Postgraduate Award scholarship to Karen Bell.
References
Avise, J.C., Walker, D., Johns, G.C., 1998. Speciation durations and
Pleistocene effects on vertebrate phylogeography. Proc. R. Soc.
Lond. B 265, 1707–1719.
Baehr, M., 1995. Revision of Philipis (Coleoptera: Carabidae: Bembi-
diinae), a genus of arboreal tachyine beetles from the rainforests of
Eastern Australia: taxonomy, phylogeny and biogeography. Mem.
Qld. Mus. 38, 315–381.
Baker, R.H., DeSalle, R., 1997. Multiple sources of character
information and the phylogeny of Hawaiian Drosophilids. Syst.
Biol. 46, 654–673.
Bouchard, P., Brooks, D.R., Yeates, D.K., in press. J. Biogeogr.
Mosaic Macroevolution in Australian Wet Tropics Arthropods:
Community Assemblage by Taxon Pulses.
Brooks, D.R., van Veller, M.G.P., McLennan, D.A., 2001. How to do
BPA, really. J. Biogeogr. 28, 345–358.
Brown, J.H., 1988. Species Diversity. In: Myers, A.A., Giller, P.S.
(Eds.), Analytical Biogeography. Chapman & Hall, London, pp.
57–89.
Bush, M.B., 1994. Amazonian speciation: a necessarily complex
model. J. Biogeogr. 21, 5–17.
Colinvaux, P.A., 1987. Amazon diversity in the light of the paleoeco-
logical record. Quat. Sci. Rev. 6, 93–114.
Farrell, B.D., 1998. Inordinate fondness explained: why are there so
many beetles? Science 281, 555–559.
Fischer, A.G., 1960. Latitudinal variation in organic diversity.
Evolution 14, 64–81.
Givnish, T.J., 1999. On the causes of gradients in tropical tree
diversity. J. Ecol. 87, 193–210.
Hill, C.J., 1996. Habitat specificity and food preferences of an
assemblage of tropical Australian dung beetles. J. Trop. Ecol. 12,
449–460.
Hugall, A., Moritz, C., Moussalli, A., Stanisic, J., 2002. Reconcil-
ing paleodistribution models and comparative phylogeography
in the Wet Tropics rainforest land snail Gnarosophia bellenden-
kerensis (Brazier 1875). Proc. Natl. Acad. Sci. USA 99, 6112–
6117.
Joseph, L., Moritz, C., 1993. Phylogeny and historical ecology of
eastern Australian scrubwrens Sericornis spp.: evidence from
mitochondrial DNA. Mol. Ecol. 2, 161–170.
Joseph, L., Moritz, C., Hugall, A., 1995. Molecular support for
vicariance as a source of diversity in rainforest. Proc. R. Soc. Lond.
B 260, 177–182.
Kershaw, A.P., 1994. Pleistocene vegetation of the humid tropics of
northeastern Queensland, Australia. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 109, 399–412.
Kikkawa, J., Monteith, G.B., Ingram, G., 1981. Cape York Peninsula:
major region of faunal interchange. In: Keast, A. (Ed.), Ecological
Biogeography of Australia. W. Junk, The Hague, pp. 1695–1742.
Liebherr, J.K., 1986. Comparison of genetic variation in two carabid
beetles (Coleoptera) of differing vagility. Ann. Ent. Soc. Am. 79,
424–433.
 K.L. Bell et al. / Molecular Phylogenetics and Evolution 31 (2004) 741–753
Matthews, E.G., 1972. A revision of the scarabaeine dung beetles of
Australia I. Tribe Onthophagini. Aust. J. Zool., Suppl. Ser. 9,
3–330.
Matthews, E.G., 1974. A revision of the scarabaeine dung beetles of
Australia. II. Tribe Scarabaeini. Aust. J. Zool., Suppl. Ser. 24,
1–211.
Matthews, E.G., 1976. A revision of the scarabaeine dung beetles of
Australia III, Tribe Coprini. Aust. J. Zool., Suppl. Ser. 38, 1–52.
Moritz, C., Richardson, K.S., Ferrier, S., Monteith, G.B., Stanisic, J.,
Williams, S.E., Whiffin, T., 2001. Biogeographical concordance
and efficiency of taxon indicators for establishing conservation
priority in a tropical rainforest biota. Proc. R. Soc. Lond. B. Bio
268, 1875–1881.
Nix, H., 1991. Biogeography: pattern and process. In: Nix, H.A.,
Switzer, M.A. (Eds.), Rainforest Animals: Atlas of Vertebrates
Endemic to AustraliaÕs Wet Tropics. Australian National Parks
and Wildlife Service, Canberra, pp. 11–39.
Pope, L.C., Estoup, A., Moritz, C., 2000. Phylogeography and
population structure of an ecotonal marsupial, Bettongia tropica,
determined using mtDNA and microsatellites. Mol. Ecol. 9, 2041–
2053.
Reid, C.A.M., Storey, R.I., 2000. Revision of the Dung Beetle Genus
Temnoplectron Westwood (Coleoptera: Scarabaeidae: Scarabaeini).
Mem. Qld. Mus. 46, 253–297.
Roff, D.A., 1994. The evolution of flightlessness: is history important?
Evol. Ecol. 8, 639–657.
View publication stats
753
Schneider, C., Moritz, C., 1999. Rainforest refugia and evolution in
AustraliaÕs Wet Tropics. Proc. R. Soc. Lond. B. 266, 191–196.
Schneider, C.J., Cunningham, M., Moritz, C., 1998. Comparative
phylogeography and the history of endemic vertebrates in the Wet
Tropics rainforests of Australia. Mol. Ecol. 7, 487–498.
Schneider, C.J., Smith, T.B., Larison, B., Moritz, C., 1999. A test of
alternative models of diversification in tropical rainforests: ecolog-
ical gradients vs. rainforest refugia. Proc. Natl. Acad. Sci. USA 96,
13869–13873.
Shimodaira, H., Hasegawa, M., 1999. Multiple comparisons of log-
likelihoods with applications to phylogenetic inference. Mol. Biol.
Evol. 16, 1114–1116.
Simon, C., Frati, F., Beckenback, A., Crespi, B., Liu, H., Flook, P.,
1994. Evolution, weighting and phylogenetic utility of mitochon-
drial gene sequences and a compilation of conserved polymerase
chain reaction primers. Ent. Soc. Am. 87, 651–701.
Storey, R.I., Monteith, G.B., 2000. Five new species of Aptenocan-
thon Matthews (Coleoptera: Scarabaeidae: Scarabaienae) from
tropical Australia, with notes on distribution. Mem. Qld. Mus.
46, 349–358.
Williams, S.E., Pearson, R.G., Walsh, P.J., 1996. Distributions and
biodiversity of the terrestrial vertebrates of AustraliaÕs Wet Tropics:
a review of current knowledge. Pacific Conserv. Biol. 2, 327–362.
Yeates, D.K., Bouchard, P., Monteith, G.B., 2002. Patterns and levels
of endemism in the Australian Wet Tropics rainforest: evidence
from flightless insects. Invert. Systematics 16, 605–619.